Point Management Logical Name Operation

Point management logical names operate as follows:

BSM_PTM_APPQ

If:

The internal threshold value for messages queued to a client is reached (possibly specified by BSM_PTM_APPQ),

Then:

Point Management first checks the approximate period of time since communication occurred with that application.

BSM_PTM_AQ_OF_DELAY

If:

The period of time exceeds the number of seconds specified by BSM_PTM_AQ_OF_DELAY

Then:

Point Management drops messages.

If:

Communication has occurred within the allowed period of time,

Then:

Point Management begins watching for continued growth, by keeping track of the number of messages a client has consumed compared to the number of messages being queued for the client.

BSM_PTM_AQ_CNT4DROP

If:

Point Management finds that growth has occurred in the number of periods specified by BSM_PTM_AQ_CNT4DROP,

Then:

Point Management will start dropping records. Note that these periods are not required to be time consecutive, that is, growth might be noted for three time consecutive periods, no growth for 2 periods, and then growth for another three periods. It is when the maximum number of periods is exceeded that dropping will occur.

If:

The system merely has encountered a burst,

Then:

It is expected that client applications will consume queued messages, and the internal lists will drop below the threshold.

BSM_PTM_DCQ

If:

If Point Manger receives a larger volume of messages from a devcom  than set by BSM_PTM_DCQ.

Then:

Increments of the System Sentry (performance) counter recording devcom queue overflows will periodically log  messages identifying Device communications occurrence queue overflow - <device>.

When messages drop below that threshold, a reset for the periods of growth count occurs. Counting, therefore, starts over the next time the threshold is exceeded.

More information

Point Management logical names.

Base system logical names.